The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various shell commands to perform repository analysis and language auditing.
Evidence: Use of git status -sb, rg -n '^#' README.md, ls -1, git ls-files, wc -l, and make docs-sync-check across SKILL.md and modules/language-audit.md to identify project structure and verify documentation status.
Context: These commands are diagnostic and used to gather information about the repository state.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external web searches and local files, creating a surface where malicious instructions embedded in that data could influence the agent.
Ingestion points: The skill reads existing README.md files, repository manifest files (e.g., package.json, Cargo.toml), and external web search results for exemplar documentation (modules/exemplar-research.md).
Boundary markers: The instructions do not specify the use of clear delimiters or guardrail prompts to isolate the external research content from the agent's instructional context.
Capability inventory: The skill can execute shell commands (git, rg, ls, wc, make), write to the filesystem (README.md, /tmp/readme-research.md), and invoke specialized tools like scribe and sanctum.
Sanitization: There is no mention of sanitizing or filtering external content retrieved from web searches before it is processed by the agent to generate new documentation.

update-readme